Complex diseases such as diabetes, cancer, cardiovascular disease, autism and schizophrenia cause an enormous burden of morbidity and mortality. Although these diseases clearly have genetic underpinnings, we have a poor understanding of these genetic factors. Recent data have led to a model in which patients carry many rare mutations of small effect that affect a vast number of genes and additively result in a specific complex disease. However, an alternative model posits that patients carry both a few rare or common mutations of small effect as well as a mutation in a strong genetic modifier. The modifier mutation enhances the penetrance of the otherwise small-effect additive variants; evidence from autism and schizophrenia studies implicates chromatin remodeling genes in buffering of other genetic variants. Because these models cannot be rigorously tested by further genetic analysis in humans - yet the principles underlying the genetics of complex traits are likely universal - we will test them in a simple organism with powerful genetics, the yeast Saccharomyces cerevisiae. The yeast mating pathway is a model complex trait, and it uses signaling components that are highly conserved in mammals and other organisms. In Aim 1, we use a technology that we recently developed to identify thousands of small-effect variants in five genes, as well as variants in these genes that epistatically interact with the known strong genetic modifier Hsp90 or have gene-by-environment interactions with any of four distinct stresses. In Aim 2, we test the two disease models by assaying over a million combinations of small-effect variants, determining whether either perturbation of the strong genetic modifier Hsp90 or a selected stress greatly increases the number of cases of severe pathway disruption. In Aim 3, we leverage our unique experimental framework to test 15 well-supported candidate genes of diverse function, including chromatin remodeling, for their potential to act as strong genetic modifiers analogous to Hsp90. These candidate genes have close human orthologs, with mutations in all but two of these orthologs implicated in complex diseases. This analysis will enable us to compare variants buffered by different molecular mechanisms and to identify targets for future studies in humans. Taken together, the results of this proposal will yield a detailed description of: (1) the contribution of thousands of small-effect variants within a gene and across pathway genes to a complex trait; (2) the frequency, identity and effect size of variants that interact with Hsp90 and/or environmental factors; and (3) genes that act as strong genetic modifiers. Results that argue for a role of strong genetic modifiers in complex disease will have consequences for developing diagnostics, preventative measures such as life-style changes and early detection, and targeted therapies.